| 基于递归神经网络的加速度传感器动态特性补偿 |
| |
| 引用本文： | 刘刚,刘学仁. 基于递归神经网络的加速度传感器动态特性补偿[J]. 传感器与微系统, 2007, 26(7): 70-72 |
| |
| 作者姓名： | 刘刚 刘学仁 |
| |
| 作者单位： | 江西师范大学,物理与通信电子学院,江西,南昌,330022 |
| |
| 摘 要： | 提出了一种基于递归神经网络的加速度传感器动态性能补偿方法,利用神经网络良好的非线性映射能力,建立传感器的动态逆模型,用实际工作参数训练神经网络,实现对加速度传感器动态特性的补偿。实验结果表明:经过动态补偿后,加速度传感器的系统工作频带得以拓宽,检测信号达到稳态的时间从补偿前的7m s缩短到大约1m s,传感器的动态性能得到明显的改善。
|
| 关 键 词： | 加速度传感器 递归神经网络 动态补偿 |
| 文章编号： | 1000-9787（2007）07-0070-03 |
| 修稿时间： | 2006-12-31 |
Dynamic compensation of accelerometer based on recurrent neural network |
| |
| LIU Gang,LIU Xue-ren. Dynamic compensation of accelerometer based on recurrent neural network[J]. Transducer and Microsystem Technology, 2007, 26(7): 70-72 |
| |
| Authors: | LIU Gang LIU Xue-ren |
| |
| Abstract: | The method of dynamic compensation of accelerometer based on recurrent neural network is introduced.By use of building the sensor's inverse modle and training the network with actual parameter,the dynamic compensation of accelerometer is realized.The experiment results show that the system working band can be broaden,the time reaching to stable state can be decreased from 7 ms to 1 ms and the dynamic characteristics can be improved after dynamic compensation. |
| |
| Keywords: | accelerometer- recurrent neural network dynamic compensation |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
